Analysis of Incomplete Data in the Presence of Competing Risks :A Simulation Study

Document Type : Original Article

Author

Statistics,Mathematics and insurance department Faculty of Commerce Damnhour University Damnhour Egypt

Abstract

This research aims at assessing a number of parametric and nonparametric estimators that are used in the estimation of competing risks models in the presence of incomplete data using a “simulation study". The nonparametric estimators used to estimate the reliability function in the presence of competing risks and right censored data in the simulation study are kaplanmeier estimator (KM) and modified weighted kaplanmeier estimator (MWKM). In addition, the parametric estimators are exponential estimator and we bull estimator. The “relatively efficient "was used to compare the results of the simulation study. The findings showed that the parametric estimator using we bull distribution (incorrect distribution) is better compared to the nonparametric estimators. Moreover, the nonparametric estimator using “KM” is better compared to the nonparametric estimator using “MWKM”. Additionally, the est­imated parametric efficiency of the we bull distribution (incorrect distribution) is not enhanced as the sample size and the censored ratio are increasing. Furthermore, the efficiency of the nonparametric estimators using “KM” is improved; however it deteriorates when using “MWKM”.  

Keywords

Files

Share

How to cite

Statistics